Select query time out after 10 k iterations


#1

I’m facing timeout on basic select query (select * from bucket use keys ‘key’). Strange thing is query times out after around 10 k iterations and all the subsequent query iterations get timed out. I’m using java client 2.2.8. We have 5 nodes in the cluster with two nodes running query and index service.

Below is the environment object I have created in the code.

DefaultCouchbaseEnvironment.builder().connectTimeout(100000).autoreleaseAfter(10000)
.build();


#2

Can you test this using cbq shell?

Copying @daschl and @ingenthr.


#3

If it’s a simple command line test, you may want to raise the logging level to DEBUG/FINE and have a look at the logs. That should make the action jumping to the exception path clear.

By the way, client 2.2.8 is rather old at this stage. I don’t think there are any relevant changes, but it may make sense to diagnose on the latest release.

$ git show 2.2.8
tag 2.2.8
Tagger: Simon Baslé simon.basle@gmail.com
Date: Wed Jun 8 13:52:37 2016 +0200

Couchbase Java Client 2.2.8 GA


#4

@geraldss… I will check in shell and get back to you… @ingenthr when we started off, 2.2.8 was the latest release, that is why we proceeded with it, but what you say does make sense.


#5

The problem mentioned above (query timing out continuously after 10 k iterations) has stopped after I upgraded the sdk version to the latest(2.4.4). However, N1ql query timeouts still occur on a regular basis on buckets with over 400 k documents. Abundant RAM has been allocated. Just want to understand if these timeouts are something that is expected to happen or is there still something we can do to prevent timeouts from happening completely?